Transform Coefficient Entropy Coding by Block-Specific VLC Classes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image compression methods, such as those in H.264/AVC and VC-1, use a single entropy coding method for all block sizes within a picture, which can lead to suboptimal compression efficiency due to varying distributions of quantized transform coefficients across different block sizes.
Innovation Solution
Classify quantized transform coefficients into pre-defined classes based on block size and distribution patterns, and apply different entropy coding methods tailored to each class, such as Context Adaptive Hybrid Variable Length Coding (CAHVLC) and Hybrid Variable Length Coding (HVLC), to optimize coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single entropy coding method is used for all block sizes, then the device complexity is reduced and ease of operation is improved, but compression efficiency deteriorates due to suboptimal adaptation to varying coefficient distributions
Solution Approach 1:
The patent segments the entropy coding process by dividing transform coefficient blocks into different size categories (e.g., 4×4, 8×8, 16×16 blocks) and applying different entropy coding methods to each segment. This allows the system to optimize compression efficiency for each block size while maintaining manageable complexity through structured classification.
Solution Approach 2:
The patent applies local quality by tailoring the entropy coding method to match the local characteristics of each block size. Different coding methods (e.g., CAVLC for smaller blocks, CABAC for larger blocks) are selected based on the specific properties of each block size, ensuring optimal adaptation to the local coefficient distribution patterns.
2Loss of energy
If multiple entropy coding methods are applied to different block sizes, then compression efficiency is improved through better adaptation, but device complexity increases due to multiple coding methods
Solution Approach 1:
The patent introduces dynamics by making the selection of entropy coding methods adaptive and flexible. The system dynamically chooses appropriate coding methods based on block size characteristics, allowing the coding strategy to adapt to varying content requirements while maintaining a structured framework that prevents excessive complexity.
Solution Approach 2:
The patent applies parameter changes by varying the entropy coding method parameter (which method to use) based on the block size parameter. This systematic parameter variation allows the system to optimize bitstream length for different block sizes while maintaining clear control rules that prevent complexity from becoming unmanageable.
3Manufacturing precision
If multiple entropy coding methods are used for different block sizes, then coding precision is improved through optimized compression, but ease of operation deteriorates due to increased complexity in selecting and managing multiple methods
Solution Approach 1:
The patent applies preliminary action by pre-establishing clear classification rules and selection criteria for different entropy coding methods based on block sizes. This preliminary structuring of the coding strategy simplifies the operational complexity, as the system follows predetermined rules rather than requiring complex real-time decisions, thereby improving ease of operation while maintaining high coding precision.
4Ease of operation
If a single entropy coding method is used uniformly, then ease of operation is improved and device complexity is reduced, but manufacturing precision deteriorates due to suboptimal compression adaptation
Solution Approach 1:
The patent applies universality by creating a unified entropy coding framework that can accommodate multiple coding methods within a single standardized structure. This multi-functional approach allows different coding methods to be integrated into one cohesive system, maintaining ease of operation through standardized interfaces while achieving high compression precision through method diversity.
Data Source
AI summary
Classifying a series of quantized transform coefficients of a block of image data into one of a pre-defined plurality of classes for entropy coding, and entropy coding the block. A class is defined by at least the size of the block and typically but not necessarily one or more other factors. The classified series is coded by one of a set of pre-defined entropy coding methods, e.g., variable length coding methods for the pre-defined classes.